Re: Rosen, again!
DON MIKULECKY (MIKULECKY@VCUVAX.BITNET)
Mon, 13 Feb 1995 16:21:49 -0400
Don Mikulecky, MCV/VCU, Mikulecky@gems.vcu.edu
P. S. on my reply to Cliff:
I've been doing a little study on my own into recursive
functions and the issue of computability (remember, I'm just
a physiologist, not a computer scientist) and it seems to me that this
in itself should make the subject interesting and relevant, given the
discussions ongoing in this net. The issue of computability is closely
linked to that of Goedel's undecidability, and all that can of worms.
Rosen has consistantly pointed to the syntax/semantics issue as a
root distinction between simple machines and complex systems. Others
do the same and in less systematic ways. By the way, Anticipatory Systems
has a good explanation of why category theory is a help in such a discussion.
Rosen (wisely I think, but you may disagree) restricts his discussion of
complexity to mathematics, initially. The topic he focuses on is
the attempt at formalization that Goedel blew out of the water.
He shows that category theory was developed to do mathematics on
mathematics, so to speak. In our discussion, he used it in a broader
context, to show that it extends beyond math to all complex systems.
Once again, I am puzzled at why Turchin's comment could have so much
weight in this context. I hope I can turn us toward the question
of what makes a complex system and what makes a living system what it
is. I think that focusing on one or another detail in a presentation
as long in developing as this is to loose the forest for the trees.
P.P.S. In ANTICIPARTORY SYSTEMS, Rosen quotes Norbert Weiner
as he shows how to simulate a phototrophic organism with a machine.
The example is neat, because Weiner's simulatiom works, but misses the
entire point of the adaptive (final cause) nature of phototropism.
He also has a nice discussion of feedback, negative and positive,
and reminds us that controll theory is generally a linear systems theory
and that cybernetics seems to like the idea of using simple machines as
simulations of complex systems as much as the rest of the reductionist
world! Maybe this is too hard on cybernetics? I wonder.
Best wishes, Don Mikulecky